Petals.ai provides Iris, an AI‑driven subject matter expert bot that ingests legacy source code and extracts business logic into structured documentation, functional specifications, and test cases. The conversational interface lets developers, testers, and analysts query processes, regulations, and inventory rules in real time, reducing reliance on hard‑to‑replace human experts and accelerating legacy system modernization.
Funding
Funding not disclosed
Founders
Product
Problem
Legacy enterprise applications are often built on outdated technologies with sparse documentation, making it difficult for teams to understand business logic, maintain the system, or add new features. This knowledge gap creates bottlenecks during modernization projects and forces reliance on hard‑to‑replace subject matter experts.
Solution
Petals.ai offers Iris, an AI‑driven subject matter expert bot that ingests legacy source code, interacts with existing experts, and extracts business knowledge into structured documentation. Iris generates functional specifications and test cases that reflect the current system’s behavior, enabling developers to validate refactoring, rebuilding, or retention efforts. The bot also provides a conversational Q&A interface for testers, analysts, and new staff to query business processes, regulatory impacts, and inventory rules in real time. By automating knowledge capture and test case creation, Iris reduces dependence on human experts, shortens onboarding, and supplies reliable inputs for AI‑assisted code generation tools.
Target Audience
Primary customers are enterprise software teams and business analysts responsible for modernizing legacy systems in supply chain, retail, and manufacturing organizations.
Features
- Connects directly to source code repositories to analyze millions of lines of legacy code and derive business logic
- Interactive knowledge acquisition from domain experts through guided conversations, enriching the AI model with company‑specific rules
- Generates objective, measurable functional test cases covering 50‑75% of existing functionality for validation and reuse
- Real‑time Q&A bot that answers detailed process, regulatory, and inventory questions using extracted code and documentation knowledge
- Supports modernization strategies focused on refactor, rebuild, and retain, providing test artifacts that feed AI code‑generation platforms (e.g., GitHub Copilot, Amazon CodeWhisperer)
- Reduces expert involvement by over 50% and accelerates new employee onboarding from weeks to days